The primary goals of the proposed supplementary effort are three-fold (supplement to LM-7085 (P41) : (1) To provide an ongoing mechanism to test and evolve force fields. To focus on the force field issues, a conformational search method will be identified and agreed upon by the interest group. The CASP infrastructure will be used to provide testing mechanism. In addition, a large variety of incorrect ('decoy') structures will be collected, characterized, and provided for benchmarking. The prediction community will be requested to contribute these decoys. A pilot server for decoys has already been set up as part of the CASP_D experiment. Evaluations will be made available to the community. Participants performing individual benchmarking will be asked to make available the results of their tests as well. In this way, it will be possible to objectively assess the strengths and weaknesses of present force fields, and to identify the most effective ways of improving them. (2) To provide an ongoing mechanism to test and evolve conformational search methods. Many groups have developed methods of searching the conformational space for the functional conformation. A major complicating factor is that most new searches also use new force fields, and it has been difficult to separate the effects of the two components. To focus on the search issues, a series of force fields will be used in testing, one at a time. Details will be agreed upon by the interest group. The CASP infrastructure will be used to provide testing mechanism. As with potentials, examples of search problems will be sought from the prediction community, and the results obtained on this collection with particular search methods made public. (3) To provide a communication system that will allow full sharing of the results of the project within the physics-based prediction community, participation in directing the project, and effective discussion of the scientific significance of the results. None of the objectives can be achieved unless they have the support and confidence of the community. We believe that we are well placed to foster such support as a result of the CASP structure prediction experiment, but each component will require careful consultation.